BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Graph Neural Networks for skillful weather forecasting - Dr Ferran
  Alet - Research Scientist\, Google DeepMind
DTSTART:20240214T150500Z
DTEND:20240214T155500Z
UID:TALK211114@talks.cam.ac.uk
CONTACT:Ben Karniely
DESCRIPTION:The dynamics of weather systems are among the most complex phy
 sical phenomena on Earth\, and each day\, countless decisions depend on ac
 curate weather forecasts\, from deciding whether to wear a jacket or to fl
 ee a dangerous storm. Until recently\, the dominant approach for weather f
 orecasting was “numerical weather prediction” (NWP)\, which involves s
 olving the governing equations of weather using supercomputers. In the las
 t year\, deep learning methods have surpassed NWPs at deterministic global
  weather forecasting and are showing promising results in probabilistic fo
 recasting as well.\n\nThis talk will focus on the use of Graph Neural Netw
 orks(GNNs) for weather forecasting. First\, we will give a brief overview 
 on how GNNs are related to Finite Element Methods used in traditional mode
 ls\, leveraging some of their useful inductive biases(Alet et al. ‘19). 
 Then\, we will delve into two recent works on GNNs for weather forecasting
 : 1. GraphCast (Lam et al. ‘23)\, which used GNNs for state-of-the-art d
 eterministic weather forecasting at 0.25 degrees of resolution\, as well a
 s predicting tropical cyclones tracks and 2. GenCast (Price et al. ‘23)\
 , which combined ideas from GraphCast and diffusion models for probabilist
 ic weather forecasts at 1 degree better than the best physics model.\n\nSo
 urces:\n\n- Graph Element Networks: adaptive\, structured computation and 
 memory (Alet et al.\, ICML ‘19)\nhttps://arxiv.org/abs/1904.09019\n\n- G
 raphCast: learning skillful medium-range weather forecasting (Lam et al.\,
  Science ‘23)\nhttps://arxiv.org/pdf/2212.12794.pdf\n\n- GenCast: diffus
 ion-based ensemble forecasting for medium-range weather (Price et al.\, ar
 xiv ‘23)\nhttps://arxiv.org/abs/2312.15796\n\nLink to join virtually: ht
 tps://cam-ac-uk.zoom.us/j/81322468305\n\nA recording of this talk is avail
 able at the following link: https://www.cl.cam.ac.uk/seminars/wednesday/vi
 deo/
LOCATION:Lecture Theatre 1\, Computer Laboratory\, William Gates Building
END:VEVENT
END:VCALENDAR